This cohort study investigates the association of allostatic load with racial and rural-urban disparities in survival among patients with stage I to III breast cancer.
Key Points
Question
Does allostatic load help explain racial and rural disparities in breast cancer survival?
Findings
In this cohort study of 3069 patients with stage I to III breast cancer, higher allostatic load was associated with an increased risk of mortality. Decomposition analyses suggested that the contribution of allostatic load to racial and rural disparities in survival was not statistically significant.
Meaning
This study found that allostatic load was associated with worse breast cancer survival, highlighting the need for interventions that address chronic stress and promote health equity.
Abstract
Importance
Racial and geographic disparities in breast cancer survival persist. Biological consequences of chronic stress, measured by allostatic load (AL), may contribute to these disparities, but their role in breast cancer prognosis is not well understood.
Objective
To evaluate the association between AL and overall survival among patients with breast cancer and assess the contribution of AL to racial and rural disparities in survival.
Design, Setting, and Participants
This cohort study used data from an institutional electronic health record and cancer registry at the University of Virginia Comprehensive Cancer Center. Participants were women diagnosed with stage I to III breast cancer between 2014 and 2024. Data were analyzed from February 2025 through April 2025.
Exposures
Allostatic load was derived from 14 clinical biomarkers and medication history and categorized as low (score ≤3) or high (score >3).
Main Outcomes and Measures
The primary outcome was overall survival. Cox proportional hazards models were used to assess the association with overall survival. Blinder-Oaxaca decomposition was performed to assess the extent to which AL explained disparities in survival by race and rurality.
Results
Among 3069 patients with stage I to III breast cancer followed up for a median (IQR) of 55.4 (30.1-83.3) months (353 Black [11.5%], 2530 White [82.4%], and 178 other race [5.8%]; 1565 aged ≤66 years [51.0%] and 1504 aged >66 years [49.0%]), most individuals were postmenopausal (1523 patients [49.6%]) and married (1766 patients [57.5%]). There was a higher mean (SD) AL score in patients who were older (4.39 [1.68] for ages >66 years vs 3.84 [1.68] for ages ≤66 years), Black (4.54 [1.81] vs 4.07 [1.68] for White patients), rural residents (4.25 [1.72] vs 4.08 [1.70] for urban residents), uninsured (4.35 [1.55]) or on public insurance (eg, 4.33 [1.93] for Medicaid vs 3.70 [1.56] for private insurance), and unemployed (4.14 [1.77]) or retired (4.39 [1.65] vs 3.63 [1.60] for employed) and who reported no alcohol use (4.33 [1.74] vs 3.94 [1.66] for reported use) or a history of tobacco use (4.20 [1.69] vs 4.05 [1.72] for no reported use). High AL score (>3) was associated with an increased risk of mortality (hazard ratios [HRs] ranged from 1.26; 95% CI, 1.04 to 1.54 for model 5 [adjustment for demographics, social economic status, lifestyle factors, and clinical factors] to 1.53; 95% CI, 1.26 to 1.86 for model 1 [crude model]) compared with low AL score (≤3). Stratified analyses indicated a larger HR in the association between AL score and mortality among rural Black patients (HR per 1-unit increase in AL, 3.33; 95% CI, 1.27-8.77). In decomposition analyses, the total explained portion accounted for −0.0708 of the racial disparity, with AL score accounting for −0.0100 (95% CI, −0.0219 to 0.0018), or 14.2% of the explained portion, while the total explained portion of the geographic disparity was −0.0127, with AL score accounting for −0.0043 (95% CI, −0.0107 to 0.0021), or 34.3% of the explained rural disparity, although estimates were not statistically significant.
Conclusions and Relevance
In this study, high AL was independently associated with worse overall survival. The contribution of AL to the observed racial and rural disparities did not reach statistical significance.
Introduction
Breast cancer is the most frequently diagnosed cancer and a leading cause of cancer mortality among women in the US.1 While advancements in screening and treatment have improved overall survival, persistent disparities remain in breast cancer outcomes, particularly by race and geographic location.2,3 Black women, for instance, are more likely to be diagnosed at later stages and have significantly lower survival rates compared with White women despite similar or lower incidence rates.4 Likewise, women living in rural areas often face higher mortality risk compared with their urban counterparts due in part to differences in health care infrastructure, access to care, and social determinants of health.5,6,7
These disparities underscore the need to examine upstream factors that may contribute to differential breast cancer outcomes across populations. One such factor is allostatic load (AL), a measure of the cumulative physiological burden imposed by chronic psychosocial and environmental stress.8,9 AL reflects the dysregulation of multiple biological systems, including metabolic, cardiovascular, and inflammatory pathways, that can result from prolonged exposure to stressors.10 Increased AL has been associated with a range of adverse health outcomes, including premature mortality and increased risk of chronic disease.9,10 In the context of breast cancer, chronic stress and the biological sequelae captured by AL may influence tumor biology,11 treatment response,12,13 and, ultimately, survival.14,15
Importantly, AL is not uniformly distributed across populations.16 Structural inequalities, including racism, economic disadvantage, and geographic isolation, contribute to greater stress exposure and physiological wear and tear among marginalized groups.17 Black individuals and those living in rural areas are disproportionately burdened by factors such as poverty, discrimination, limited access to health care, and social isolation, all of which can contribute to increased AL.18,19 These disparities in AL may help explain observed inequities in breast cancer outcomes and offer new insights into the biological embodiment of social disadvantage.
Using patients with breast cancer seen at the University of Virginia Comprehensive Cancer Center in this study, we aimed to determine whether1 pretreatment AL was associated with overall survival,2 whether this association was moderated by race and urban vs rural residence,3 and to what extent AL may explain racial and geographic disparities in breast cancer survival. By integrating biological, social, and geographic dimensions of health, this research sought to elucidate pathways through which chronic stress contributed to inequities in breast cancer survival.
Methods
Study Population
In this cohort study, 3069 female patients with breast cancer were identified from a pool of patients with breast cancer seen at the University of Virginia Comprehensive Cancer Center in the last decade (2014-2024). The selection criteria included newly diagnosed stage I to III breast cancer, no prior cancer diagnosis except nonmelanoma skin cancer, and living in the catchment area of the University of Virginia Cancer Center, which includes 3.2 million residents from a large area that includes 87 counties throughout northern, central, southside, and southwestern Virginia, as well as eastern West Virginia. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was followed in the design, analysis, and interpretation of study results. The Institutional Research Board at the University of Virginia approved this study. Informed consent was waived by the institutional review board because this was a retrospective review.
AL Construction
In this study, AL was assessed using 14 biomarkers routinely collected in clinical practice and commonly referenced in the AL literature.16,20,21 Briefly, biomarkers were selected to represent 4 physiological systems: cardiovascular (heart rate, systolic blood pressure, and diastolic blood pressure), metabolic (body mass index [calculated as weight in kilograms divided by height in meters squared] and triglyceride, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, total cholesterol, alkaline phosphatase, fasting glucose, and albumin levels), kidney (creatinine levels, estimated glomerular filtration rate, and blood urea nitrogen levels), and immune (white blood cell count) systems. In addition to these biomarkers, medication history for diabetes, cardiovascular disease, chronic kidney disease, and hypertension was incorporated into the AL score. To address missing data, multiple imputation by chained equations22 was performed to generate 10 imputed datasets. Each biomarker was dichotomized based on established clinical risk thresholds (eTable 1 in Supplement 1), assigning a score of 1 (deemed as high risk) or 0 (deemed as low risk). The AL score was computed as the sum of all positive indicators, yielding a total score ranging from 0 to 15, with higher scores reflecting greater physiological dysregulation. For analytic purposes, AL scores were further categorized into 2 groups: low AL (score ≤3) and high AL (score >3) based on the distribution of scores within the study population.
Assessment of Covariates
Sociodemographic factors included age, self-reported race (Black, White, and other), menopausal status (premenopausal, perimenopausal, postmenopausal, and other), employment status (employed, unemployed, disabled, and retired), marital status (married and other), insurance status (private, Medicaid, Medicare, military, and uninsured), tobacco use (ever and never), and alcohol consumption (ever and never). Neighborhood characteristics included area deprivation index (ADI)23 and Rural-Urban Continuum Codes (RUCC).24 ADI was dichotomized into high deprivation (≤5) and low deprivation (>5). RUCC was classified as rural (RUCC >3) or urban (RUCC ≤3). Tumor characteristics included tumor stage and tumor subtype. Tumor stage was classified according to the American Joint Committee on Cancer TNM staging system into stages I, II, and III. Tumors were classified into subtypes, including by estrogen receptor (ER) status and triple-negative breast cancer status, with triple-negative breast cancer tumors defined as lacking expression of ER, progesterone receptor, and human epidermal growth factor receptor 2. Additionally, clinical treatment variables, including chemotherapy and radiation therapy, were each classified as yes or no based on treatment receipt.
Outcomes
The primary outcome was overall survival, defined as the time from the date of breast cancer diagnosis to the date of death from any cause. For participants who were alive at the end of the study period, survival time was censored at the date of last known contact or at study completion (September 23, 2024), whichever occurred first. The median (IQR) follow-up for study participants was 55.4 (30.1-83.3) months overall. The median (IQR) follow-up was 60.7 (34.7-86.3) months for patients who were alive and 31.2 (13.6-51.4) months for patients who died.
Statistical Analysis
Mean differences in AL were evaluated across demographic, lifestyle, clinical, tumor, and socioeconomic characteristics. Poisson regression assessed the joint association of race and geographic location with AL. Associations between AL and breast cancer mortality were examined using Cox proportional hazards models, with sequential adjustments. Model 2 included demographics (age, race, and menopausal status). Model 3 added socioeconomic status (employment, marital, and insurance status). Model 4 included lifestyle factors (tobacco and alcohol use). Model 5 added clinical characteristics (tumor stage and subtype, chemotherapy, and radiation). Model 6 was fully adjusted, incorporating individual- and neighborhood-level factors (area deprivation index and RUCC). Stratified analyses further explored AL-mortality associations within subgroups. To evaluate the contribution of AL to racial and geographic disparities in 3-year breast cancer survival, Blinder-Oaxaca decomposition was applied using separate models for race (Black vs White) and geography (rural vs urban). Models were specified as linear probability regressions and adjusted for key covariates, including age, tumor stage, alcohol use, and triple-negative status. Pooled regression coefficients were used as weights, and 1000 bootstrap replications generated 95% CIs and P values for total and variable-specific contributions. All analyses were conducted using R statistical software version 4.3.0 (R Project for Statistical Computing), with 2-sided tests and significance defined as P < .05. Data were analyzed from February 2025 through April 2025.
Results
The study included 3069 patients with breast cancer (353 Black [11.5%], 2530 White [82.4%], and 178 other race [5.8%]; 1565 aged ≤66 years [51.0%] and 1504 aged >66 years [49.0%]), with a mean (SD) AL score of 4.12 (1.70). Most patients were postmenopausal (1523 patients [49.6%]) and married (1766 patients [57.5%]). AL levels varied significantly across demographic subgroups (Table 1). Older patients had a higher mean (SD) AL compared with younger patients (4.39 [1.68] for ages >66 years vs 3.84 [1.68] for ages ≤66 years; P < .001). Racial disparities were evident, with Black patients exhibiting the highest mean (SD) AL score (4.54 [1.81]) vs White patients (4.07 [1.68]) and other racial groups (3.89 [1.66]; P < .001). Women who were postmenopausal (4.14 [1.67]) and those with other menopausal status (eg, hysterectomy: 4.21 [1.76]) had higher mean (SD) AL scores than women who were premenopausal (3.45 [1.69]; P < .001). Socioeconomic status factors were also associated AL, with employed individuals (3.63 [1.60]) showing lower mean (SD) AL scores than unemployed (4.14 [1.77]) or retired (4.39 [1.65]) individuals (P < .001) and uninsured (4.35 [1.55]) patients or those with public insurance (eg, 4.33 [1.93] for Medicaid) showing higher mean (SD) AL scores than patients with private insurance (3.70 [1.56]; P < .001). Mean (SD) AL score was also higher women in rural areas than those in urban area (4.25 [1.72] vs 4.08 [1.70]; P = .04). Lifestyle factors, including tobacco use and alcohol consumption, were also associated with AL levels. Patients who reported ever drinking (3.94 [1.66] vs 4.33 [1.74] for never drinking; P < .001) and those who reported never smoking (4.05 [1.72] vs 4.20 [1.69] for ever smoking; P = .01) had lower mean (SD) AL scores than their counterparts. In terms of clinical variables, mean (SD) AL score was higher in women diagnosed with ER-positive breast cancer than those with ER-negative breast cancer (4.14 [1.72] vs 3.95 [1.59]; P = .02). Notably, patients who died during follow-up had significantly higher mean (SD) AL scores than survivors (4.57 [1.83] vs 4.02 [1.66]; P < .001). Next, we explored the joint association of racial and geographic differences with AL among patients with breast cancer (eTable 2 in Supplement 1). Black patients in rural areas had the highest mean AL score (4.70; 95% CI, 4.10-5.30), with an increased relative ratio (1.11; 95% CI, 1.01-1.22) compared with White urban patients. Black urban patients also exhibited increased AL scores (relative ratio, 1.07; 95% CI, 1.01-1.14) compared with White urban patients.
Table 1. AL Levels by Patient Characteristics.
| Characteristic | Patients, No. (%) | AL score, mean (SD) | P value |
|---|---|---|---|
| Overall | 3069 (100) | 4.12 (1.70) | NA |
| Age group, y | |||
| ≤66 | 1565 (50.99) | 3.84 (1.68) | <.001 |
| >66 | 1504 (49.01) | 4.39 (1.68) | |
| Race | |||
| Black | 353 (11.50) | 4.54 (1.81) | <.001 |
| White | 2530 (82.44) | 4.07 (1.68) | |
| Othera | 178 (5.80) | 3.89 (1.66) | |
| Menopausal status | |||
| Premenopausal | 213 (6.94) | 3.45 (1.69) | <.001 |
| Perimenopausal | 52 (1.69) | 3.52 (1.55) | |
| Postmenopausal | 1523 (49.63) | 4.14 (1.67) | |
| Otherb | 878 (28.61) | 4.21 (1.76) | |
| Marital status | |||
| Married | 1766 (57.54) | 3.99 (1.70) | <.001 |
| Otherc | 1297 (42.26) | 4.28 (1.70) | |
| Insurance status | |||
| Private | 1699 (55.36) | 3.70 (1.56) | <.001 |
| Medicare | 750 (24.44) | 4.42 (1.67) | |
| Medicaid | 138 (4.50) | 4.33 (1.93) | |
| Military | 49 (1.60) | 4.27 (1.81) | |
| Uninsured | 433 (14.11) | 4.35 (1.55) | |
| Employment status | |||
| Employed | 964 (31.41) | 3.63 (1.60) | <.001 |
| Unemployed | 372 (12.12) | 4.14 (1.77) | |
| Disabled | 190 (6.19) | 4.41 (1.68) | |
| Retired | 1271 (41.41) | 4.39 (1.65) | |
| ADI within the state of Virginia | |||
| ≤5 | 1633 (53.21) | 4.12 (1.70) | .71 |
| >5 | 1436 (46.79) | 4.10 (1.71) | |
| RUCC | |||
| Urban | 2371 (77.26) | 4.08 (1.70) | .04 |
| Rural | 660 (21.51) | 4.25 (1.72) | |
| Alcohol consumption | |||
| Never | 1293 (42.13) | 4.33 (1.74) | <.001 |
| Ever | 1697 (55.29) | 3.94 (1.66) | |
| Tobacco use | |||
| Never | 1762 (57.41) | 4.05 (1.72) | .01 |
| Ever | 1248 (40.66) | 4.20 (1.69) | |
| Tumor stage | |||
| I | 1843 (60.05) | 4.10 (1.69) | .60 |
| II | 844 (27.50) | 4.11 (1.73) | |
| III | 358 (11.64) | 4.20 (1.73) | |
| ER status | |||
| Negative | 500 (16.29) | 3.95 (1.59) | .02 |
| Positive | 2569 (83.71) | 4.14 (1.72) | |
| Triple negative | |||
| No | 2789 (90.88) | 4.13 (1.72) | .13 |
| Yes | 280 (9.12) | 3.98 (1.57) | |
| Chemotherapy treatment | |||
| Yes | 1104 (35.97) | 4.13 (1.71) | .74 |
| No | 1965 (64.03) | 4.11 (1.70) | |
| Radiation treatment | |||
| Yes | 1589 (51.78) | 4.11 (1.73) | .93 |
| No | 1480 (48.22) | 4.12 (1.67) | |
| Death | |||
| Yes | 494 (16.10) | 4.57 (1.83) | <.001 |
| No | 2537 (82.67) | 4.02 (1.66) |
Abbreviations: ADI, area deprivation index; AL, allostatic load; ER, estrogen receptor; NA, not applicable; RUCC, Rural-Urban Continuum Codes.
Other race includes Asian, Native Hawaiian, or Other Pacific Islander and multiracial (ie, >1 race category) individuals.
Other menopausal status includes ablation, hysterectomy, and other.
Other marital status includes divorced, legally separated, single, widowed, other, and significant other.
We next assessed the association between AL and breast cancer overall survival. Kaplan-Meier curves (Figure 1) demonstrated a survival disadvantage for patients with high AL (>3) vs low AL (≤3) scores (log-rank P < .001). The divergence in survival probabilities emerged early and widened over time. Higher AL was consistently associated with increased mortality risk across all adjusted models (Table 2). In the crude model (model 1), AL score as a continuous variable was associated with an increased hazard of death (HR per 1-unit increase in AL score, 1.19; 95% CI, 1.13-1.25). This association persisted after sequential adjustment for demographics (model 2), socioeconomic status (model 3), lifestyle factors (model 4), clinical variables (model 5), and neighborhood characteristics (model 6: HR, 1.14; 95% CI, 1.08-1.21; P < .001). In a sensitivity analysis among only 2332 patients with complete biomarker data, the HR for death per 1-unit increase in AL score was 1.16 (95% CI, 1.08-1.24; P < .001). When AL score was dichotomized as high (>3) vs low (≤3), high AL was associated with increased mortality risk across models, with HRs ranging from 1.26 (95% CI, 1.04-1.54) for model 5 to 1.53 (95% CI, 1.26-1.86) for model 1. In model 6, the HR was 1.27 (95% CI, 1.04-1.55). Additionally, we conducted a sensitivity analysis using AL quartiles. Results from these quartile-based analyses were consistent with our primary findings (eTable 3 in Supplement 1).
Figure 1. Kaplan-Meier Survival Estimates for the Association Between Allostatic Load Score and Breast Cancer Mortality.
Allostatic load was categorized as low (score ≤3) or high (score >3).
Table 2. Association Between AL and Overall Survival.
| AL score | Death, HR (95% CI) | |||||
|---|---|---|---|---|---|---|
| Model 1a | Model 2b | Model 3c | Model 4d | Model 5e | Model 6f | |
| Continuous, per 1-unit increase | 1.19 (1.13-1.25) | 1.17 (1.11-1.23) | 1.14 (1.08-1.20) | 1.13 (1.07-1.19) | 1.14 (1.08-1.20) | 1.14 (1.08-1.21) |
| Categorical | ||||||
| Low (≤3) | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| High (>3) | 1.53 (1.26-1.86) | 1.42 (1.17-1.74) | 1.29 (1.06-1.58) | 1.28 (1.05-1.57) | 1.26 (1.04-1.54) | 1.27 (1.04-1.55) |
Abbreviation: AL, allostatic load; HR, hazard ratio.
Crude model.
Model 1 plus demographics (including age, race, and menopausal status).
Model 2 plus social economic status (including employment status, marital status, and insurance status).
Model 3 plus lifestyle factors (including alcohol use and tobacco use).
Model 4 plus clinical factors (including tumor stage, estrogen receptor status, triple-negative status, chemotherapy, and radiation treatment).
Model 5 plus neighborhood variables (including area deprivation index and Rural-Urban Continuum Codes).
We further investigated whether the risk association between AL and breast cancer survival differed by race and urban vs rural residence. The associated increase in risk was higher among Black (HR, 1.18; 95% CI, 1.01-1.39) than White (HR, 1.15; 95% CI, 1.08-1.21) patients with breast cancer. For urban vs rural residence, the increase in risk was significant among patients with urban residence (HR, 1.16; 95% CI, 1.09-1.23) but not among those with rural residence (HR, 1.10; 95% CI, 0.99-1.22). When we combined race and urban vs rural residence (Figure 2), racial and geographic factors interacted with the risk association between and AL and breast cancer survival. Per 1-unit increase in AL, the hazard of death increased among urban White (HR, 1.16; 95% CI, 1.08-1.25), urban Black (HR, 1.26; 95% CI, 1.04,-1.53), and rural White (HR, 1.09; 95% CI, 0.98-1.22) patients. Notably, rural Black women exhibited a substantially increased hazard of death (HR, 3.33; 95% CI, 1.27-8.77) per unit increase in AL score. We further assessed interactions among race, rural residence, and AL score; however, none reached statistical significance.
Figure 2. Association Between Allostatic Load and Breast Cancer Mortality by Race and Rurality.
HR indicates hazard ratio.
Finally, we performed Blinder-Oaxaca decomposition to evaluate to what extent AL may explain racial and geographic disparities in breast cancer survival (Table 3). Among Black and White patients, the total disparity was −0.0148, reflecting worse outcomes among Black patients. The explained portion accounted for −0.0708 of the disparity, while the unexplained component (0.0560) offset part of the gap. Allostatic load alone explained −0.0100 (95% CI, −0.0219 to 0.0018), or 14.2% of the explained disparity; although its contribution was not statistically significant (P = .11), it ranked as the third-largest contributor after tumor stage (−0.0270; 95% CI, −0.0470 to −0.0070 [38.1%]; P = .01) and triple-negative breast cancer (−0.0223; 95% CI, −0.0423 to −0.0033 [31.5%]; P = .02). In the rural-urban comparison, the total disparity was larger (−0.0304), with observed covariates accounting for −0.0127 of the disparity (41.8%) and −0.0177 of the disparity (58.2%) attributed to unexplained factors. AL emerged as one of the leading contributors to the explained rural-urban disparity, accounting for −0.0043 (95% CI, −0.0107 to 0.0021), or 34.3% of the explained disparity. Although this contribution was not statistically significant (P = .17), it exceeded that of tumor stage (−0.0042; 95% CI, −0.0164 to −0.0011 [33.2%]; P = .02) and age (−0.0021; 95% CI, −0.0083 to 0.0041 [16.6%]; P = .51).
Table 3. Blinder-Oaxaca Decomposition Analyses of Role of AL in Mortality Disparities.
| Variable | Estimated mean | Proportion of explained disparity | Proportion of total explained, % | P value | |
|---|---|---|---|---|---|
| Comparison group | Reference group | ||||
| Racial disparity a | |||||
| AL | 4.49 | 4.07 | −0.0100 (−0.0219 to 0.0018) | 14.20 | .11 |
| Age | 64.77 | 67.04 | −0.0004 (−0.0103 to 0.0094) | 0.60 | .93 |
| Stage | 0.19 | 0.11 | −0.0270 (−0.0470 to −0.0070) | 38.10 | .01 |
| Alcohol use | 0.38 | 0.58 | −0.0150 (−0.0337 to 0.0044) | 21.10 | .10 |
| TNBC | 0.21 | 0.08 | −0.0223 (−0.0423 to −0.0033) | 31.50 | .02 |
| Rurality | 0.28 | 0.22 | 0.0040 (−0.0036 to 0.0116) | −5.70 | .33 |
| Total explained | NA | NA | −0.0708 | 100 | NA |
| Unexplained | NA | NA | 0.0560 | NA | NA |
| Total disparity | NA | NA | −0.0148 | NA | NA |
| Geographic disparity b | |||||
| AL | 4.27 | 4.08 | −0.0043 (−0.0107 to 0.0021) | 34.25 | .17 |
| Age | 67.31 | 66.58 | −0.0021 (−0.0083 to 0.0041) | 16.63 | .51 |
| Stage | 0.13 | 0.12 | −0.0042 (−0.0164 to −0.0011) | 33.19 | .02 |
| Alcohol use | 0.46 | 0.58 | −0.0054 (−0.0164 to 0.0057) | 42.40 | .35 |
| TNBC | 0.11 | 0.10 | −0.0009 (−0.0034 to 0.0014) | 6.79 | .39 |
| Race | 0.16 | 0.12 | 0.0042 (−0.0036 to 0.0120) | −33.26 | .33 |
| Total explained | NA | NA | −0.0127 | 100 | NA |
| Unexplained | NA | NA | −0.0177 | NA | NA |
| Total disparity | NA | NA | −0.0304 | NA | NA |
Abbreviations: AL, allostatic load; NA, not applicable; TNBC, triple-negative breast cancer.
Racial disparity is between Black (comparison group) and White (reference group) patients.
Geographic disparity is between rural (comparison group) and urban (reference group) patients.
Discussion
In this cohort study of more than 3000 patients with breast cancer treated at the University of Virginia Comprehensive Cancer Center, increased AL, a marker of cumulative physiological stress, was associated with decreased overall survival. However, the magnitude of HRs in this association varied across population subgroups. Black women and those residing in rural areas had significantly higher AL scores than their White and urban counterparts. While increased AL was associated with worse survival across the cohort, HRs were highest among Black women regardless of geographic location. Rural Black women had the highest AL levels, suggesting greater cumulative stress exposure, and rural Black women exhibited the highest HR in the association between AL and mortality risk. These patterns indicate that both race and geographic context may interact with the association between AL and survival, potentially reflecting differences in structural stressors, access to care, or tumor biology. Further research is warranted to disentangle these intersecting factors and their roles in breast cancer disparities.
We found that AL varied significantly across demographic and socioeconomic subgroups. Black patients, those living in rural areas, and individuals with lower socioeconomic status exhibited higher AL scores, consistent with existing literature showing that marginalized populations are disproportionately exposed to chronic stressors.18,19,25 These stressors, rooted in systemic racism, poverty, and geographic isolation, contribute to the dysregulation of multiple physiological systems. Our observation that Black women, particularly those living in rural areas, had the highest AL scores underscores the intersectional nature of structural disadvantage and its physiological consequences.
The survival analysis revealed that higher AL score was associated with increased mortality risk. Patients with high AL scores had significantly worse survival outcomes, with each unit increase in AL corresponding to an increase in the hazard of death across adjusted models, with HRs ranging from 1.13 to 1.19. This dose-response pattern is consistent with the hypothesis that AL captures cumulative biological wear and tear that may impair the body’s capacity to cope with cancer and its treatment. Our results are consistent with previous reports by Obeng-Gyasi et al,14,15 in which they found a similar association between increased AL and all-cause mortality in patients with breast cancer. The association between AL and poor outcomes may reflect stress-related disruptions in immune function, inflammation, endocrine regulation, and metabolic balance, all of which can influence tumor progression, treatment resistance, and recovery.26,27 Notably, the association between AL and mortality varied by race and geographic location. Stratified and joint effect analyses revealed that rural Black women experienced the highest mortality risk associated with increased AL, followed by urban Black and urban White women. Rural Black women exhibited a markedly increased risk per 1-unit increase in AL of death (HR, 3.33), compared with a smaller increase among urban White women (HR, 1.16), emphasizing the complex and multifactorial nature of racial and geographic disparities in breast cancer outcomes.28
In this study, we applied the Blinder-Oaxaca decomposition method to examine factors contributing to racial and rural disparities in breast cancer outcomes, with a particular focus on AL as a potential explanatory factor. While AL did not reach statistical significance, it accounted for a meaningful portion of the explained disparity: 14.2% in racial comparisons and 34.3% in rural-urban comparisons. These contributions were on par with or exceeded those of more commonly studied factors, such as age and tumor stage. However, these findings should be interpreted cautiously. The lack of statistical significance may, in part, reflect limited sample sizes; only 11.5% of the study population identified as Black, and 21.5% of patients resided in rural areas, thereby reducing power to detect robust associations. Nonetheless, these results are consistent with prior research associating chronic stress with adverse breast cancer outcomes, especially in socially and structurally marginalized populations.29,30 Although exploratory, these findings point to the potential biological relevance of AL in cancer disparities and underscore the need for further research in more diverse and adequately powered cohorts.
Strengths and Limitations
Despite the strengths of this study, including its large, regionally diverse sample and comprehensive adjustment for confounders, several limitations must be acknowledged. AL was measured using available clinical biomarkers, which represent a clinically feasible AL measure for retrospective electronic health record–based studies and may not capture the full breadth of physiological stress responses, such as neuroendocrine markers. Residual confounding due to unmeasured stressors, such as discrimination, trauma, or caregiving burden, may have also influenced our estimates. In addition, while multiple imputation was used to address missing data, measurement error in self-reported variables could introduce bias. To address this concern, we conducted a sensitivity analysis restricted to patients with breast cancer with complete biomarker data. The analysis revealed a statistically significant increase in the hazard of death per 1-unit increase in AL score (HR, 1.16; 95% CI, 1.08-1.24, P < .001) after adjusting for all covariates. Additionally, as an observational study, this study cannot establish causality. We cannot rule out reverse causality (eg, worse cancer prognosis increasing AL). Thus, our findings should be interpreted within the context of a regional cancer center that may not reflect other health care settings.
Conclusions
This cohort study found that increased AL was associated with shorter breast cancer mortality, and our findings may contribute to understanding persistent racial and rural disparities in survival outcomes. While contributions of AL to these disparities did not reach statistical significance, observed patterns suggest that chronic stress and physiological dysregulation may play a supportive role in shaping inequities. These findings underscore the potential value of integrating biological measures of stress into cancer disparities research, alongside social determinants of health. Future studies with larger and more diverse populations are needed to clarify the role of AL in cancer outcomes. In the meantime, addressing chronic stress and promoting supportive care strategies may hold promise for improving survivorship, especially among Black women and rural patients who may experience disproportionate stress burdens.
eTable 1. Distribution and high-risk thresholds of individual biomarkers for allostatic load
eTable 2. AL levels by racial and geographic differences among patients with breast cancer
eTable 3. Sensitivity analysis to assess the association between AL and overall survival among patients with breast cancer
Data Sharing Statement
References
- 1.Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229-263. doi: 10.3322/caac.21834 [DOI] [PubMed] [Google Scholar]
- 2.Fan Q, Yao XA, Han X. Spatial variation and disparity in female breast cancer relative survival in the United States. Cancer. 2021;127(21):4006-4014. doi: 10.1002/cncr.33801 [DOI] [PubMed] [Google Scholar]
- 3.Ellington TD, Henley SJ, Wilson RJ, Miller JW, Wu M, Richardson LC. Trends in breast cancer mortality by race/ethnicity, age, and US census region, United States—1999-2020. Cancer. 2023;129(1):32-38. doi: 10.1002/cncr.34503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Primm KM, Zhao H, Hernandez DC, Chang S. A contemporary analysis of racial and ethnic disparities in diagnosis of early-stage breast cancer and stage-specific survival by molecular subtype. Cancer Epidemiol Biomarkers Prev. 2022;31(6):1185-1194. doi: 10.1158/1055-9965.EPI-22-0020 [DOI] [PubMed] [Google Scholar]
- 5.LeBlanc G, Lee I, Carretta H, Luo Y, Sinha D, Rust G. Rural-urban differences in breast cancer stage at diagnosis. Womens Health Rep (New Rochelle). 2022;3(1):207-214. doi: 10.1089/whr.2021.0082 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chandak A, Nayar P, Lin G. Rural-urban disparities in access to breast cancer screening: a spatial clustering analysis. J Rural Health. 2019;35(2):229-235. doi: 10.1111/jrh.12308 [DOI] [PubMed] [Google Scholar]
- 7.Obeng-Gyasi S, Obeng-Gyasi B, Tarver W. Breast cancer disparities and the impact of geography. Surg Oncol Clin N Am. 2022;31(1):81-90. doi: 10.1016/j.soc.2021.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rodriquez EJ, Kim EN, Sumner AE, Nápoles AM, Pérez-Stable EJ. Allostatic load: importance, markers, and score determination in minority and disparity populations. J Urban Health. 2019;96(suppl 1):3-11. doi: 10.1007/s11524-019-00345-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Guidi J, Lucente M, Sonino N, Fava GA. Allostatic load and its impact on health: a systematic review. Psychother Psychosom. 2021;90(1):11-27. doi: 10.1159/000510696 [DOI] [PubMed] [Google Scholar]
- 10.Pfaltz MC, Schnyder U. Allostatic load and allostatic overload: preventive and clinical implications. Psychother Psychosom. 2023;92(5):279-282. doi: 10.1159/000534340 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Liu HM, Ma LL, Li C, et al. The molecular mechanism of chronic stress affecting the occurrence and development of breast cancer and potential drug therapy. Transl Oncol. 2022;15(1):101281. doi: 10.1016/j.tranon.2021.101281 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gosain R, Gage-Bouchard E, Ambrosone C, Repasky E, Gandhi S. Stress reduction strategies in breast cancer: review of pharmacologic and non-pharmacologic based strategies. Semin Immunopathol. 2020;42(6):719-734. doi: 10.1007/s00281-020-00815-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chen JC, Elsaid MI, Handley D, et al. Allostatic load as a predictor of postoperative complications in patients with breast cancer. NPJ Breast Cancer. 2024;10(1):44. doi: 10.1038/s41523-024-00654-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Obeng-Gyasi S, Elsaid MI, Lu Y, et al. Association of allostatic load with all-cause mortality in patients with breast cancer. JAMA Netw Open. 2023;6(5):e2313989. doi: 10.1001/jamanetworkopen.2023.13989 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen JC, Elsaid MI, Handley D, et al. Association between neighborhood opportunity, allostatic load, and all-cause mortality in patients with breast cancer. J Clin Oncol. 2024;42(15):1788-1798. doi: 10.1200/JCO.23.00907 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Shen J, Fuemmeler BF, Guan Y, Zhao H. Association of allostatic load and all cancer risk in the SWAN cohort. Cancers (Basel). 2022;14(13):3044. doi: 10.3390/cancers14133044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Clark EC, Cranston E, Polin T, et al. Structural interventions that affect racial inequities and their impact on population health outcomes: a systematic review. BMC Public Health. 2022;22(1):2162. doi: 10.1186/s12889-022-14603-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Esdaille AR, Kuete NK, Anyaeche VI, Kalemoglu E, Kucuk O. The interplay between structural inequality, allostatic load, inflammation, and cancer in Black Americans: a narrative review. Cancers (Basel). 2024;16(17):3023. doi: 10.3390/cancers16173023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Duru OK, Harawa NT, Kermah D, Norris KC. Allostatic load burden and racial disparities in mortality. J Natl Med Assoc. 2012;104(1-2):89-95. doi: 10.1016/S0027-9684(15)30120-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shen J, Fuemmeler BF, Sheppard VB, et al. Neighborhood disadvantage and biological aging biomarkers among breast cancer patients. Sci Rep. 2022;12(1):11006. doi: 10.1038/s41598-022-15260-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Guan Y, Shen J, Lu J, Fuemmeler BF, Shock LS, Zhao H. Association between allostatic load and breast cancer risk: a cohort study. Breast Cancer Res. 2023;25(1):155. doi: 10.1186/s13058-023-01754-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20(1):40-49. doi: 10.1002/mpr.329 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kind AJH, Buckingham WR. Making neighborhood-disadvantage metrics accessible—the Neighborhood Atlas. N Engl J Med. 2018;378(26):2456-2458. doi: 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Economic Research Service . Rural-urban continuum codes. US Department of Agriculture. Accessed July 21, 2025. https://www.ers.usda.gov/data-products/rural-urban-continuum-codes
- 25.Hébert JR, Braun KL, Kaholokula JK, Armstead CA, Burch JB, Thompson B. Considering the role of stress in populations of high-risk, underserved community networks program centers. Prog Community Health Partnersh. 2015;9(0)(suppl):71-82. doi: 10.1353/cpr.2015.0028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Liu YZ, Wang YX, Jiang CL. Inflammation: the common pathway of stress-related diseases. Front Hum Neurosci. 2017;11:316. doi: 10.3389/fnhum.2017.00316 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Acabchuk RL, Kamath J, Salamone JD, Johnson BT. Stress and chronic illness: the inflammatory pathway. Soc Sci Med. 2017;185:166-170. doi: 10.1016/j.socscimed.2017.04.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hirko KA, Rocque G, Reasor E, et al. The impact of race and ethnicity in breast cancer-disparities and implications for precision oncology. BMC Med. 2022;20(1):72. doi: 10.1186/s12916-022-02260-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sánchez-Díaz CT, Strayhorn S, Tejeda S, Vijayasiri G, Rauscher GH, Molina Y. What mediates the racial/ethnic disparity in psychosocial stress among breast cancer patients? Cancer Causes Control. 2021;32(4):357-367. doi: 10.1007/s10552-021-01392-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sanogo F, Jefferson M, Beard TA, et al. Social and clinical drivers of stress responses in African American breast cancer survivors. Sci Rep. 2024;14(1):19729. doi: 10.1038/s41598-024-70841-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Distribution and high-risk thresholds of individual biomarkers for allostatic load
eTable 2. AL levels by racial and geographic differences among patients with breast cancer
eTable 3. Sensitivity analysis to assess the association between AL and overall survival among patients with breast cancer
Data Sharing Statement
